Modeling the evolution of inter-hospital referral networks: An agent-based modeling approach

Tao Yang¹Jie Fan²Wenbin Gu¹Jing Qin¹Bing Deng³Wenwen Deng^3*Wei Wang^1*

• ¹Chongqing Medical University College of Public Health, Chongqing, China
• ²Chongqing Medical and Pharmaceutical College, Chongqing, China
• ³Chongqing Public Health Medical Center, Chongqing, China

In many countries, especially large developing nations like China, hierarchical healthcare systems face persistent structural challenges such as unbalanced resource allocation and inefficient referral networks. To capture the impact of individual behavior on the evolution of healthcare system structures, this study develops an agent-based modeling (ABM) approach to simulate the dynamic formation of inter-hospital referral networks. The model introduces a strategic game-theoretic mechanism among patients, where each agent makes referral decisions based on multiple factors, including medical benefits, commuting costs, and limited information, under resource constraints. Through these decentralized decisions, a directed and weighted referral network emerges, reflecting realistic behavioral patterns. Based on the simulated network, topological analysis is performed to identify key hospitals and design targeted resource allocation strategies. The results show that tertiary hospitals exhibit high centrality, indicating a structural 'Matthew effect'. In addition, compared to simply strengthening core hospitals, the referral bottleneck mitigation strategy can reduce patient loss, while also decreasing the overall referral frequency within the system. Based on these findings, the study proposes a multilevel reform framework spanning operational, institutional, and strategic dimensions to support the restructuring of hierarchical healthcare systems through game-theoretic and optimization approaches.